|From the paper mentioned below
That was an imaginary example, but here are two real ones. In the last week or so we have been bombarded with articles with headlines like 'Astronomers find alien megastructure', where in reality they have found evidence for a group of smaller objects in orbit around a star, which could have many natural explanations. Yes, in principle it could also be caused by a vast structure - but this is extremely unlikely.
Even worse, a few years ago we saw the headlines 'Star Wars lightsabers finally invented,' and 'Scientists Finally Invent Real, Working Lightsabers.' I love that 'finally' - I mean, come on, scientists, get your act together, George Lucas had them nearly 40 years ago. Those are both actual headlines. But, no, scientists did not invent 'real, working lightsabers'. What they did was produce a genuinely interesting effect where two photons briefly linked together. That was it. (Admittedly in this case, a scientist involved rather foolishly used the 'lightsaber' word in a press release and started it all.)
This kind of vastly over-inflated headline to draw the reader in seems self-defeating. The result is that after a few disappointments you don't trust the headlines. You know you are going to be a let down. But at least in many cases the body text of the article makes it clear that the headline was over-the-top. However sometimes the problem extends right through the text, and that's where the reporting is not just a matter of an exaggerated title but of irresponsible representation of the facts.
The piece that started me on this was one dating back to 2013, dramatically titled Physicists create world's first multiverse of universes in the lab. Being used to eye-catching headlines, I assumed that the statement would be an exaggeration. For instance, maybe they constructed a multiverse model in a computer program - which isn't really 'in the lab'. But, no, this was a physical experiment involving a metamaterial, an electromagnet and polarised light. And throughout the article it was clear that this was the real thing. The researchers, we are told, 'created multiple universes.' Wow, and furthermore, phew! As the article says, and who can argue, 'this is the first ever time that new universes have been created in a laboratory setting.'
I naturally wanted to find out more, and the article helpfully gives you a link to the paper. There I was confronted with this in the abstract: 'Extraordinary light rays propagating inside a hyperbolic metamaterial look similar to particle world lines' and 'thermal fluctuations in a ferrofluid look similar to creation and disappearance of individual Minkowski spacetimes (universes) in the cosmological multiverse.' Don't worry if you aren't familiar with all the terminology used here. The key words here are perfectly everyday English: 'look similar to.'
Now, I'm sorry, but those are very important words, which are simply not given the weight they should have in the article that caught my eye in the first place. After all, I've seen pictures of a cat on Facebook that looks similar to Hitler. But it would be a bit of a stretch to write an article saying 'cat owner discovers Hitler in her front room.' This is a slightly exaggerated illustration, I admit, but there is a world (nay, a multiverse) of difference between something 'looking similar' to something else and something being something else.
You might think I have no room to complain because the article did include a link to the paper, so anyone could do what I did and check the details - and that's great. I applaud that opportunity. But the vast majority of people won't do that. And so another science myth gets perpetrated. Those in the science community often bemoan the lack of understanding of science in the general public - but as long as we continue with this kind of exaggeration, then science writers and other journalists who pick up on science-based stories (and, yes, scientists and PR officers who misrepresent stories in press releases) need to shoulder a fair amount of the blame.